AI-first models and doing business in an AI world
AI-first companies are new players on the rise. Nowadays, delivering value through systems of intelligence provides a key success in optimizing services. Industries are changing and becoming more competitive in delivering solutions to customers. Entrepreneurs need to consider current technological trends when developing new services. Meanwhile, incumbent companies often need to remodel their entire systems to suit current trends.
And that’s exactly what AI-first companies offer.
AI-first companies was the topic of a recent episode of The Twenty Minute VC. Ash Fontana, the managing director at Zetta Venture Partners, was the guest star. Together with the host, he indulged in an enlightening discussion about the undeniable change in business models that companies are offering nowadays. The discussion covered concepts such as vertical integration in businesses and delivering customer value through predictive models using systems of intelligence.
What are AI-first companies and their characteristics?
“AI-first companies focus on building better models instead of product features; quantitative feedback instead of qualitative feedback; have machine agents running itself rather than humans [running it]; and it automatically iterates on these features rather than manually.” — Ash Fontana
AI-first companies are modern companies that take hold of the reverberating artificial intelligence phenomenon when offering their products and services. The characteristics of such a company are unique. The priority and emphasis here are on predictive models. Predictive models are structures that can forecast customer and market trends while continuously learning from new trends. They do that by utilizing large data sets of customer information, which allows executives to make more informed decisions for their businesses. In turn, it becomes easier to create meaningful customer value. AI-centered startups, as an example, take an advantageous position compared to traditional companies lacking such a model.
Every function within AI-first companies relates to improving the models. Firstly, AI-first companies build a predictive model. Next, they create a method to deliver customer-tailored products and services. Customer data is central to the creation and the resulting development of the model, as continuous feedback loops need to be in place.
In the wake of AI, contrary to what many people might think, the process still very much needs human involvement. This is where Fontana introduces the term virtuous loop — the idea that collecting the right customer data is crucial and forms the interactive component in interactive machine learning. In other words, human feedback is still at the core of the process and must be included in the loop.
AI-first companies versus traditional companies
Incumbent businesses are often associated with traditional organizations. However, it’s important to note that they’re not always synonymous. In the meantime, AI-first companies are currently establishing their place in the market, and for that reason, a lot of people view them as newcomers. Yet, given the prominence of AI, AI-first companies are at an advantage. It skips the challenging transition that traditional organizations face.
What’s this challenging transition, you ask? Well, it’s the very significance of utilizing customer data. Fontana explains that incumbents struggle to become AI-first simply because they exist before this phenomenon. They don’t have the agility that a smaller, more versatile startup has in adopting new practices revolving around predictive analytics. Consequently, these companies lack the predictive part of the business model. This prevents them from being competitive and reacting quickly to the ever-increasing speed of market changes.
As Simon Sinek would say, executives and entrepreneurs just need to ask the ‘Why’ question, as many of them do not understand their customers. Why are their customers, their customers? Answering this question will lead to a thinking that tries to understand the extra, if not deeper, layers of providing customer solutions and values.
With an AI-first principle, companies are better able to understand their customers by analyzing the right data. Following that, customized solutions that attend to the tailored needs of the customer can materialize. That’s an extra score for AI-first companies.
Systems of record and systems of intelligence
When comparing and contrasting traditional and AI-first companies, Fontana differentiates systems of record from systems of intelligence. Systems of record acquire data in a structured form as humans manually enter it. Systems of intelligence automatically acquire data in an unstructured form as machines collect it. The latter principle is becoming more and more favorable, as it supports businesses in the decision-making process. Companies are able to acquire suggestions based on their predictive models with this. In contrast, systems of record just own and record’s assets, leaving businesses with no valuable information that could optimize the workflow.
Traditional industries implementing an AI-first principle
The following are some great examples to consider when visualizing how traditional incumbents with big data libraries are implementing intelligent systems.
The stock exchange market represents one of the many traditional industries with a lot of data. With the move towards AI, the stock market has seen investing bots conquering quantitative trading. Data by JPMorgan Chase & Co. shows that 60% of all equity assets in the U.S. are now controlled by quantitative investment funds, thanks to bots. A subset of quantitative trading, known as high-frequency trading, also skyrocketed more than half (52%) of May’s stock market volume in the U.S.
Another great example is AI in healthcare. The Internet of Medical Things (IoMT) allows health professionals to better understand the routines of individuals through the use of consumer health apps and wearables, states PwC. Orbita integrates smart home devices, wearables, and home health monitors to create better awareness and care for people having chronic illnesses. Its assistive solutions are delivered through voice assistants and chatbots, which are solutions that are developed in close collaboration with nurses.
Shifting from SaaS models to systems of intelligence
The same SaaS (Software as a Service) models can’t be applied anymore. AI-first company yet again satisfies the criteria for a more successful model in the context of data-driven projects.
Customers often view the SaaS model as a cost-saving method. This is due to its ability to improve a user’s workflow by facilitating platforms that can automate tasks. But this merely generates value on an individual level with a “one-size fits all” type solution. Contrast this to AI-first companies that create a shared-benefit model where each customer is offered unique solutions because of the unique data they have.
Vertical integration and AI-first
According to Fontana, “For an AI-first company, [vertical integration] means doing anything from creating your own proprietary data, creating your own model, and delivering these predictions to customers through your own products.” Essentially, AI-first companies offer their owned customizable services to the integration of a customer’s business process.
For instance, Amazon started as a warehouse and logistics company to finally offer cloud and streaming services. Vertical integration is especially evident in Amazon Prime due to the many features it offers — from free shipping to streaming music, movies, and TV shows. More importantly, customers’ experiences in using its services are customizable and tailored to their own needs. But thanks to AI, anticipating customer purchasing behaviors is Amazon Prime’s e-commerce and shipping specialty. AI has successfully forecasted customer product demands and consequently offer faster delivery options as fast as one-hour deliveries. Effectivity and efficiency are points that Amazon Prime has scored. That quick delivery time is definitely a milestone for the world of logistics.
Now, think back to the concept of mass customization. The success of Nike By You customized sneakers was unprecedented, which led to its continual operations today. The ability to personalize and customize Nike sneakers enabled unique solutions and outcomes for the individual, and that’s exactly the concept that relates to an AI-first principle. Once a customized pair of Nike sneakers are designed and purchased, Nike goes into production and runs the vertical integration wheel from there. From design to manufacturing, to final delivery, the company has wide-ranging operations to better suit the demands of customers. If Nike were to adopt an AI-first operation for Nike By You, imagine the possibilities that it can bring.
Takeaway
Nowadays, we involve data in everything we do; this could be online and/or offline data. The challenge is in using such information and running algorithms to predict possible outcomes. Modern business models need to maximize the potential of customer data, and this could only be done with an AI-first principle.
Author: Neysa Tavianto
Feature photo by vectorpouch — www.freepik.com
AI Lab One is an artificial intelligence consulting agency. For information on what we do and what makes us tick, visit https://ailab.one/
